Vol. 28. pp. 3185 to 3189.
DIMERIZATIONS OF METHYLENECYCLOPROPANES W. R. DOLBIER, JR,* D. LOMAS, T. GAIUA, C. HARMON* and P. TARRANT Dtpartmentof Chemistry, University of Florida, Gainesville, Florida 32601 (Received br the USA 10 January 1972; Received br the UKforpublkation
3 March 1972)
Abmlmct-The tlmmal I2 + 21 dimerization of metbylenecyclopropanes, including methylenecyclopropane itself, is discussed. The effects of structuralfeatures on the ability of this -ion to Occurare discussed in terms of the probable mechanism. INTRODUCTION
SIMPLEALKENESdo not usually undergo (2 + 21 thermal dimerizations. It has generally been found that in order for such reactions to proceed to any sign&ant extent, there must be either a relatively stable diradical or dipolar intermediate in the two-step cycloaddition process, e.g. the dimerixations of acrylonitrile’ or allene,2 or a powerful &mux.Iynamic driving force for destruction of the alkene in favor of the cyclobutane, e.g. polyfluorinated systems.‘* 3 Recently, we discovered that certain methylenecyclopropanes underwent relatively facile 12 + 21 dimerizations.‘ We now report further details of this reaction and others of this type which seem to intimate that the reaction should be rather general in systems where significant non-bonded interactions are not present in the cyclobutane product. RESULTS
Initially we found that th: very reactive dichloromethylenecyclopropane (la) formed the head-to-head dimcr quantitatively, at temperatures as low as 100”. With this facile R R D-=
la, R=Cl lb, R=H lc, R=Me
R R 2a, R=Cl zb, R=H 2c, R=Me
14 n=l le. n=2 l A. P. Sloan Foundation Fellow, 1970-72. $ Participant in NSF-URP program at University of Florida, Summer, 1969.
W. R. Do~em,
JR., D. Lams,
T. GARZA,C. HARMON
and P. Tm
synthesis of the dispiroI220.127.116.11 octane system, we converted tetrachloride 2a, to the interesting hydrocarbon, dispiro[ 18.104.22.168 act-7ene (4), which proved to be remarkably stable to unimolecular thermal decomposition, but very prone to polymerization at room temperature, even in dilute solution. H
The observation of this facile dimerization led to the question of how general this type of reaction was. Soon after, Conia reported the dimerization of biscyclopropylidine (ld) which occurred at 210”,’ and we found that cyclopropylidenecyclobutane (le) also underwent thermal dimcrization with relative ease (210’, six hours). At this point we turned to the parent species (lb) and found it not nearly so reactive. Nevertheless, it also dimerized to about 20% upon heating at 240-4F for 48 hours. A limit to the generality of the dimerization process was reached, however, with the attempt to dime&e isopropylidenecyclopropane (1~). Upon heating 2,2dimethylmethylenecyclopropane (5) at temperatures ranging from 210 to 245’ for times ranging from four to eighty hours, no dimeric species could be isolated. 5 has been reported to convert quantitatively to lc upon heating to 225°.6 The product mixtures were shown to contain no isolable products boiling above 80°. Moreover NMR spectra
and GLPC traces indicated that less than 5% of dimeric, cyclopropane-containing species were present. At higher temperatures, or for longer time runs, no lc or 5 could be detected as remaining in the product mixture. DISCUSSION
is no doubt that the major driving force for the conversion of 1 -+ 2 is the ratedetermining destruction of the sp2 cyclopropane-ring carbon. Having an sp2 carbon in a cyclopropane ring apparently gives rise to a very substantial increment of strain energy of the order of 12 kcal/mole,’ the destruction of which would provide for a signif&nt thermodynamic advantage for [ 2 + 21 cycloadditions of methylenecyclopropanes both in the initial step and also in the overall process. The reactivity trend of the various methylenecyclopropanes which we have invesThere
Dimerixaticms of mcthyknccyclopropane
tigated, seems to be most consistent with a two-step mechanism involving a diradical intermediate (6). If such is the case, the initial bond is most-certainly formed at the ring carbons so as to take fuh advantage of the relief of strain, and to avoid formation of the relatively unstable cyclopropyl radical. For the reaction to be faciie, it is apparently necessary that 6 be reasonably stable, as it is for Is where R=Cl. When R =H, and primary radicals must prevail, the reaction is sluggish, but not entirely excluded. With R = Me, non-bonded interactions probably impede the cyclixation step severely enough so that other reactions compete favorably with dimerization. When the alkyl groups are ‘tied back’ as with Id and fe, the problem of non-bonded interactions is apparently somewhat abated, thus allowing a reasonably facile dimerization. It should be mentioned that if’diradical6 is involved as an intermediate; it must be short-lived, since cyclopropyl carbinyl radicals themselves have been demonstrated to convert irreversibly, and with relative ease, to ally1 carbinyl radicalsgVg While the activation energy for this process has not been determined, there can be little doubt that it is si~~c~t since both gas and liquid phase chlo~atio~ of m~hylcy~~opro~e Iead to significant amount of unrearmnged c~orom~y~cyclopro~e.*“, *l On the other hand, the activation energy for cyclization of the tetramethylene diradical should be relatively small (estimated by O’Neal and Benson to be -6 kcaUmole’*). Certainly, no dimers without cyclopropane rings were detected in our experiments. Mechanistically, then, our conclusion must be that the intermediacy of a short-lived diradical is most consistent with the data, although a concerted process cannot rigorously be excluded. While there does seem to be a steric limitation to the dimerization reaction, the process nevertheless shows promise of wide scope. Moreover, it seems probable that methylenecyclopropanes may participate generally in [ 2e 21 cycloadditions as do the ~l~uor~a~ olefins. It should be mentioned that me~yl~~ycioprop~e itself was earlier found to undergo a 12 + 21cycloaddition with tetraguoroethylene. I3This reaction could however have been attributed to the reactivity of the tetrafluoroethylene, especially since these workers could not observe any similar cycloadditions with maleic anhydride and acrylonitrile both of which are generally reactive in 12 + 21 processes. It now seems probable that la, (R=CI), should be very reactive in general [ 2 + 2) processes and that such reactions could provide reasonably simple synthetic routes to various spiro-, dispiro-, trispiro- and tetraspiro systems.
EXPERIMENTAL Analyses were determined by Atlantic Microlab, Inc., Atlanta, Ga 30308. The GLPC analyses were carried out on a Varian Aerograph Mode1 A-90-P3 gas crouch combined with a L.ceds and Northrup Model H recordcz. IR spectra were nrotdcd with a PerKin-ElmerModel 13’7spectrophotometer, KBr prism; NMR spectra on a Varian Model A-60-A spectrometer, utilizing TMS as an inter& standard; mass spe+a on a Hitachi Perk&Elmer RMU-6E mass spectrometer; and UV spectra on a Gary 15 spectrometer. ” 8Og of allene was condensed in to a 1 liter flask equipped with 2,2-DichloromethykneiyclopmecyclbpropMe. mechanical stirrer,dry ice condenser and dropping-fmmeeiwith N, inlet, containing O-4 1 pentane and 40 g NaOMe (cooled to -40“). Then 96 gcthy1trichloraacaate was added ail at once and the mixture albwed to warm slowly while stirring. At -loo, re8uxing began and at -8O the soiution turneda tar coior. The mixture was allowed to warm overnight, collecting excess allene in a dry ice trap, and -200 ml water added. The aq. layer was washed with 200 ml pentane layers combined, dried and distilled at atm pressure through a 1 fi vigreaux column. The fraction distilling between PO- 100” was collected (60% pure), the only
Dimerixations of methylenecyclopropanes
products were formed. GLPC indicated that no appreciable amount of material boiling over 80’ was produced, (other than polymeric), refro.spiro[ 22.214.171.124.2.0.2O]tetradecMe (2e). Cyclopropylidene cyclobutane l was sealed in a Pyrex tube and pyrolyzed at 2 IO0 for 4 hr. Small amounts of volatile materials were removed on the vacuum line and 2e was purified by GLPC using the carbowax (20 m) column at 170°. NMR, two sytnmetricaJ [email protected]
at 6 0.20 (4H) and 040 (4H) and a complex broad group ofpeaks between 1.3and 2.3 ppm ( I2 H). (Calc. for C,,H,,: C, 89.4; H, 10.6. Found: C, 89.33; H, 10.69%). Acknowledgemenr-We Foundation.
wish to acknowledge the partial support of this research by the National Science
I J. D. Roberts and C. M. Sharts. Org. Reactions 12. I (1962) ’ S. H. Dai and W. R. Dolbier, Jr., J. Org. C/rem. 37, 950 (1972) 3 C. R. Patrick, Tetrahedron 4, 26 (1958) ’ W. R. Dolbier, D. Lomas and P. Tarrant, J. Am. Chem. Sot. 90, 3594 (1968) ’ P. LePerchec and J. M. Coma, Tefrohedron Letters 1587 (1970) 6 R. Srinivasan and S. Boue, 1. Am. Chem. Sot. 93, 5606 (1971) ’ OK. B. Wiberg and R. A. Fenoglio,Ib&f. 90,3395 (1968);‘R. B. Turner, P. Goebel, B. J. Mallon, W. von E. Doering, J. F. Cobum, Jr. and M. Pomerantx, I&f. 90, 4315 (1968) * J. Kochi, P. J. Krusic and D. R. Eaton, Ibid. 91, 1877 (1969) 9 E. J. Huyser and J. D. Taliaferro,J. Org. Chem. 2S, 3443 (1963) lo J. D. Roberts and R. Maxur,J. Am. C/rem. Sot. 82, 1987 (1961) I1 C. Walling and P. S. Fredericks, Ibid. 84, 3326 (1962) r* H. E. O’NeaJ and S. W. Benson, J. P&s. Chum. 72, 1866 (1968) I3 B. C. Anderson, J. Org. Chem. 27, 2720 (1962) I4 H. G. Peer and A. Schors, Rec. Trov. Chim. 161, 86 (1967) I’ P. Gassman, J. Org. Chem. 29, 160 (1964) I6 P. S. Bailey, Znd. Eng. Chem. 50, 993 (1958) I’ L. Eberson, A&. Chem. Scud. 13, 40 (1959) I* R. Koster, S. Arora and P. Binger, Angew. Chem. Internot. Edit. 8, 205 (1969) re W. J. Bailey and C. R. Pfeifer, 1. Org. Chem. 20, 95 (1955) 2o W. Rahman and H. G. Kuivila, Ibid. 31, 772 (1966) ” J. Vincent, A. Bezaguet and M. Bertrand, Bull. Chbn. Sot. Fr. 3550 (1967)
We thank Prof. Marcel Betrand, Universite D’Aix Marseille for a sample of this compound.*’